Preprint

SWE-Pruner Pro: The Coder LLM Already Knows What to Prune

Yuhang Wang Yuling Shi Shaoqiu Zhang J. Liang Shilin He Siyu Ye Yuting Chen Kai Cai Xiaodong Gu

Jul 2026 · 2 citations · 61 references

Computer Science

TL;DR

SWE-Pruner Pro is proposed, which prunes tool outputs directly inside the agent, with a small head turns the agent's own internal representations into a keep-or-prune label for each line, with a length-aware embedding keyed to each tool output's line count.

Abstract

Pruning long context for coding agents has been a vital technology for efficient context management. While existing context pruning methods such as SWE-Pruner realize this by attaching a separate code classifier, we find the agent itself encodes internal representations indicating the relevance of code context when reading tool output. Based on this finding, we propose SWE-Pruner Pro, which prunes tool outputs directly inside the agent. Concretely, a small head turns the agent's own internal representations into a keep-or-prune label for each line, with a length-aware embedding keyed to each tool output's line count. Across two open-weight backbones and four multi-turn benchmarks, SWE-Pruner Pro saves up to 39% of prompt and completion tokens while preserving task quality, with bounded inference overhead. Notably, on MiMo-V2-Flash SWE-Pruner Pro additionally raises the SWE-Bench Verified resolve rate by +3.8% and the long-context Oolong accuracy by +2.2 points.

arXiv Semantic Scholar

View source

Similar papers

Preprint Aug 2026

CodeGrep: An RL-Trained Retrieval Agent for LLM Coding Agents

This study introduces CodeGrep, a 14B retrieval agent trained end-to-end with GRPO to issue multi-turn parallel grep, glob, and read tool calls and return candidate files to a frozen downstream coding agent, and applies the efficiency signal at the advantage layer rather than the reward layer to reduce KL drift and tra...

Wu-Ya Chen, Yihao Yang, Yang Cao et al. · 2 citations

Preprint Aug 2026

Paritok-4B: Intent-Conditioned Context Compression for Coding Agents

This work presents Paritok-4B, a 4B LoRA compressor for coding-agent trajectories built on two commitments, and distil a gpt-4.1-mini teacher over 67,074 real OpenHands trajectories into 40,606 validated examples and fine-tune Qwen3-4B.

Jiayu Shi, Lu Chen · 0 citations

#natural language process... Jul 2026

ORCA-bench: How Ready Are Language Model Agents for Oncall?

The ORCA-bench benchmark is introduced, a benchmark that puts general-purpose coding agents in a production-fidelity oncall setting and is a lower bound on the engineering investment required before frontier coding agents can be safely entrusted with production reliability.

Albert Gong, Kyuseong Choi, Abhineet Agarwal et al. · arXiv.org · 1 citation

Jul 2026

CORVUS: Context Optimization and Reduction Via Underlying Synchronization for LLM Coding Agents

LLM coding agents operate by constructing trajectories that accumulate reasoning, tool calls, and results to enable multi-step decision-making. However, the conventional append-only trajectory architecture found in practice tightly couples file-read actions with their observations, capturing snapshots that become perma...

Mingwei Zheng, David OBrien, Siwei Cui et al. · arXiv.org · 2 citations

#small language model Preprint Aug 2026

Compress and Forget: bitsandbytes Quantization Amplifies Proactive Interference in LLMs

The results suggest that bitsandbytes 4-bit quantization can impose an additional cost on applications relying on long, updatable, semantically dense contexts, even when aggregate benchmark accuracy appears largely unaffected.

Shayan Shahrabi-Farahani, D. Rahmati · 1 citation

Jul 2026

Function-Aware Fill-in-the-Middle as Mid-Training for Coding Agent Foundation Models

Beyond in-domain gains, mid-training mitigates the capability erosion that agentic post-training otherwise inflicts on non-agent coding and non-coding tool-use benchmarks (tau-bench, BFCL): although the mid-training corpus contains Python code only, the function-call inductive bias survives post-training and yields con...

Yubo Wang, Jiarong Liang, Yuxuan Zhang et al. · arXiv.org · 1 citation